Data from the analysis pipeline is written to disk in HDF5 (.h5
) format.
The number of results written to a single .h5
file is capped at 5,000.
If you are simulating a larger number of SNe, or have allocated multiple I/O
processes at runtime, then results will be written to multiple files.
The data model for each of these files is identical.
Data is written to multiple tables within each HDF5 file. The path of each table is provided below along with a summary of the corresponding data.
Table Path |
Description |
---|---|
|
A human readable message describing the status of each simulation. |
|
The parameters used as model inputs when simulating each supernova. |
|
The simulated photometry of a given supernova. |
|
The parameters recovered from fitting a model to each supernova. |
|
The covariance matrix determined when fitting a model to a given SN. |
The schema of the message
table is listed below.
Every simulation has an entry in the message table regardless of whether
the simulation for a given SNe was successful.
Column Name |
Data Type |
Description |
---|---|---|
|
String |
The ID of each simulated supernova. |
|
Bool |
Whether the simulation/fitting process executed successfully. |
|
String |
Status message describing the status of each simulation. |
Simulation parameters are appended to the simulation/params
for every single simulation
regardless of the success of the simulation.
Column Name |
Data Type |
Description |
---|---|---|
|
String |
The ID of each simulated supernova. |
|
Float |
The value of the parameter used in the simulation. |
The simulation/<ID>
tables provide the simulated light curve for each supernova.
Important
The simulation/<ID>
tables are only created if writing light-curves to
disk is enabled via the appropriate command line arguments.
Column Name |
Data Type |
Description |
---|---|---|
|
String |
The name of the bandpass for the simulated observation. |
|
Float |
The simulated observed flux through a given band. |
|
Float |
The error in |
|
Float |
The photometric zero point of the observation. |
|
Float |
The name of the zero point system used when simulating flux. |
|
Integer |
Either 0 (non-detection), 4096 (detection), or 6144 (first trigger). |
Entries in the fitting/params
table are only provided for simulations that result in successful fits.
Any fits that fail to converge are not written to the table.
Precalculated B-band magnitudes are determined using the bessellb
filter built-in to
sncosmo and the Betoule+ 2014 cosmology.
Column Name |
Data Type |
Description |
---|---|---|
|
String |
The ID of each simulated supernova. |
|
Float |
The value of the parameter recovered from the fit. |
|
Float |
The estimated error in the fitted parameter. |
|
Float |
The chisquared of the fit result. |
|
Float |
Number of degrees of freedom in the fit. |
|
Float |
Apparent b-band magnitude of the fitted model at peak. |
|
Float |
Absolute b-band magnitude of the fitted model at peak. |
Column names in the fitting/covariance/<ID>
tables act as labels for the rows and columns of the covariance matrix.
Column Name |
Data Type |
Description |
---|---|---|
|
Float |
A column of the covariance matrix. |